Allow configuring the model being used

xhjkl / well

A command-line utility to chat with your codebase

50 stars 3 forks source link

Allow configuring the model being used #9

Closed IgnisDa closed 3 months ago

IgnisDa commented 3 months ago

Right now, well uses gpt-4o for all requests. I would prefer to use gpt-3.5-turbo since I have a very big codebase and this model has a higher TPM.

IgnisDa commented 3 months ago

OPENAI_MODEL

xhjkl commented 3 months ago

Perhaps I should make it clearer in the readme, thanks for pointing this one out.